Health Effects of Cyclones: A Systematic Review and Meta-Analysis of Epidemiological Studies

Background: More intense cyclones are expected in the future as a result of climate change. A comprehensive review is urgently needed to summarize and update the evidence on the health effects of cyclones. Objectives: We aimed to provide a systematic review with meta-analysis of current evidence on the risks of all reported health outcomes related to cyclones and to identify research gaps and make recommendations for further research. Methods: We systematically searched five electronic databases (MEDLINE, Embase, PubMed, Scopus, and Web of Science) for relevant studies in English published before 21 December 2022. Following the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) guidelines, we developed inclusion criteria, screened the literature, and included epidemiological studies with a quantitative risk assessment of any mortality or morbidity-related outcomes associated with cyclone exposures. We extracted key data and assessed study quality for these studies and applied meta-analyses to quantify the overall effect estimate and the heterogeneity of comparable studies. Results: In total, 71 studies from eight countries (the United States, China, India, Japan, the Philippines, South Korea, Australia, Brazil), mostly the United States, were included in the review. These studies investigated the all-cause and cause-specific mortality, as well as morbidity related to injury, cardiovascular diseases (CVDs), respiratory diseases, infectious diseases, mental disorders, adverse birth outcomes, cancer, diabetes, and other outcomes (e.g., suicide rates, gender-based violence). Studies mostly included only one high-amplitude cyclone (cyclones with a Saffir–Simpson category of 4 or 5, i.e., Hurricanes Katrina or Sandy) and focused on mental disorders morbidity and all-cause mortality and hospitalizations. Consistently elevated risks of overall mental health morbidity, post-traumatic stress disorder (PTSD), as well as all-cause mortality or hospitalizations, were found to be associated with cyclones. However, the results for other outcomes were generally mixed or limited. A statistically significant overall relative risk of 1.09 [95% confidence interval (CI): 1.04, 1.13], 1.18 (95% CI: 1.12, 1.25), 1.15 (95% CI: 1.13, 1.18), 1.26 (95% CI: 1.05, 1.50) was observed for all-cause mortality, all-cause hospitalizations, respiratory disease, and chronic obstructive pulmonary disease hospitalizations, respectively, after cyclone exposures, whereas no statistically significant risks were identified for diabetes mortality, heart disease mortality, and preterm birth. High between-study heterogeneity was observed. Conclusions: There is generally consistent evidence supporting the notion that high-amplitude cyclones could significantly increase risks of mental disorders, especially for PTSD, as well as mortality and hospitalizations, but the evidence for other health outcomes, such as chronic diseases (e.g., CVDs, cancer, diabetes), and adverse birth outcomes remains limited or inconsistent. More studies with rigorous exposure assessment, of larger spatial and temporal scales, and using advanced modeling strategy are warranted in the future, especially for those small cyclone-prone countries or regions with low and middle incomes. https://doi.org/10.1289/EHP12158


Key Criteria
Detection bias, exposure assessment Can we be confident in the exposure characterization?

Low
There is direct evidence that exposure assessment involved representative and reliable measurements of cyclone exposure and a low risk of exposure misclassification (e.g., assessed exposure at individual level, accounted for the persistent and time-varying cyclone exposures) Probably low There is indirect evidence that the exposure assessment involved representative and reliable measurements of cyclone exposure and a low risk of exposure misclassification

Probably high
There is indirect evidence that the cyclone exposure assessment involved representative and reliable measurements, but could introduce a high risk of exposure misclassification (e.g., exposure assessment was based on a static point-in-time estimate like cyclone hit date that did not account for the cyclone end date, nor consider the persistent or time-varying cyclone exposures) OR There is insufficient information provided about the exposure assessment to judge the validity and reliability, but no evidence for concern about the method used.

High
There is direct evidence that poorly reliable and representative measurements were used to assess cyclone exposure Detection bias, outcome assessment Can we be confident in the outcome assessment?

Low
There is direct evidence that outcome data are from a reliable data source or defined based on standard diagnosis criteria (e.g., International Classification System [ICD] code) OR Studies provide evidence of quality assurance of outcome data.

Probably low
There is indirect evidence that outcome was assessed or defined using acceptable methods OR It is deemed that the outcome assessment methods used would not appreciably bias results (e.g., objectively measured and quality controlled).

Probably high
Outcome was not assessed based on standard diagnosis criteria and there is evidence that suggests the existence of misclassification bias (e.g., objectively measured but less of quality control procedure of measurement) OR There is insufficient information provided about the outcome assessment to judge the validity and reliability, but no evidence for concern about the method used.

High
Outcome was obtained or defined based on self-reports (parents, family) and data collected, criteria developed by the researcher OR There is direct evidence that suggests the high risk of outcome misclassification bias.

Confounding bias
Did the study design or analysis sufficiently account for important confounding variables?

Low
Study accounted for all important confounders which were measured consistently (e.g., age, gender, race/ethnicity, education level, household income, health status for cross-sectional, cohort and case-control studies; time trend, seasonality, day of week, public holiday, variation in expected number of outcomes for time-series study; social-economic status, region, variation in expected number of the outcome).

Probably low
Study accounted for most of confounders AND is not expected to introduce bias.
Probably high Study accounted for some but not all confounders AND is expected to introduce bias.

High
Study did not account for potential confounders OR were inappropriately measured.

Selection bias
Did selection of study participants result in appropriate comparison groups?

Low
The descriptions of the studied population were sufficiently detailed to support the assertion that risk of selection effects was minimal (e.g., study participants in different exposure levels and with all outcomes had equal opportunity to be included in the study).

Probably low
There is insufficient information about population selection to permit a judgment of low risk of bias, but there is indirect evidence that suggests low risk of bias (e.g., study participants in different exposure levels may not have equal opportunity to be in the study).
Probably high There is insufficient information about population selection to permit a judgment of high risk of bias, but there is indirect evidence that suggests high risk of bias (e.g., participants in all exposure levels did not have equal opportunity to be in the study; but not to the extent that seriously bias the effect estimates).

High
There were indications from descriptions of the studied population of high risk of bias (study only included designated high-risk participants, and participants in all exposure levels did not have equal opportunity to be in the study, to the extent that effect estimates were seriously biased).

Attrition/exclusion bias
Were outcome data complete without attrition or exclusion from analysis?

Low
No missing data irrelevant to the true study outcome and no missing outcome data Probably low Though not sufficient information available to evaluate the incomplete data's risk accurately, there was indirect evidence indicating a low risk of bias. Probably high Inadequate information provided to determine whether a risk was high about incomplete data, but there was indirect evidence to suggest a high risk. High Direct evidence to suggest that the missing data on outcomes is relevant to the true study outcome

Selective reporting bias
Did the study report all measured outcomes? Low The study reported findings on all pre-specified outcomes Probably low Inadequate information provided to determine whether a risk of selective outcome was low, but there was indirect evidence to suggest that study was not selectively reported Probably high Inadequate information provided to determine whether a risk of selective outcome was low, but there was indirect evidence to suggest that study was selectively reported

High
The study did not report findings on all pre-specified outcomes, or used methods that were not pre-specified to analyze one/more of the primary outcomes or report the outcomes/findings that were not pre-specified Conflict of interest Was there potential bias in the reporting through financial sources?

Low
No funding was received for this study from entities with a financial interest in the study outcomes. Probably low Inadequate information provided to to determine a low risk, but there was indirect evidence to suggest that the study had no financial interest Probably high Inadequate information provided to to determine a low risk, but there was indirect evidence to suggest that the study had financial interest High Support was received for this study from entities with a financial interest in the study outcomes.

Other source of bias
Bias from other sources not covered elsewhere (statistical methodological appropriateness, researcher compliance with study protocol) Low No other sources of bias Probably low Inadequate information provided to determine a low risk, but there was indirect evidence to suggest that the study had no other problems Probably high Inadequate information provided to determine a low risk, but there was indirect evidence to suggest that the study had other problems High At least one important bias detected from other sources   Probably low The exposed period for the study area was defined as 12 months after the Hurricane Michael hit date and unexposed period was defined as the 12 months before the Hurricane Michael hit date, while did account for time-varying cyclone exposures.

Low
Outcome data were derived from the local authorities and were defined based on well-established standard A separate survey was conducted before the cyclone and after the cyclone was completely over, with those surveyed after the cyclone was completely over serving as the exposed group and those surveyed before the cyclone serving as the control, but did not account for the varying exposure level